Deepfakes, Trust, and Electoral Integrity: The Emerging Threat to Indonesian Democracy

Abstract

This study examines the emerging threat of deepfakes to trust and electoral integrity in Indonesia, where digital media and social platforms increasingly mediate political communication and public opinion. As generative AI technologies enable the creation of highly realistic but fabricated audio-visual content, deepfakes pose significant risks to democratic processes by undermining the credibility of political actors, spreading misinformation, and manipulating voter perceptions. This research investigates how deepfakes are produced, circulated, and received within Indonesia’s electoral ecosystem, and assesses their potential impact on trust in institutions, media, and the electoral process. Employing a mixed-methods approach, the study combines computational detection of deepfake content, content analysis of viral misinformation cases, and qualitative interviews with journalists, election officials, digital platform experts, and civil society actors. The findings indicate that deepfakes amplify existing vulnerabilities in Indonesia’s information environment, particularly in contexts of high political polarization and limited digital literacy. Deepfake content is often strategically deployed to discredit opponents, provoke emotional reactions, and create confusion during critical electoral moments. The study also finds that public trust is eroded not only by the presence of deepfakes but by the uncertainty they generate, as audiences struggle to distinguish authentic from manipulated content. The novelty of this research lies in its context-specific analysis of deepfake threats in Indonesia’s democratic setting, offering one of the first empirical examinations of deepfake circulation in Southeast Asian elections. The contribution of the study is to develop a conceptual framework linking deepfake proliferation with trust erosion and electoral legitimacy, and to provide policy-relevant insights for strengthening digital resilience. The study concludes that protecting electoral integrity requires robust detection capabilities, platform transparency, and comprehensive public education, and recommends collaborative governance mechanisms involving state institutions, platforms, and civil society to mitigate deepfake harms.